Saturday, 25 July 2015

Portfolio Construction Part 1 – Diversification, Diworsification?

Portfolio construction is one of the most important things to get right in equity investing but also one of the hardest. While the asset selection part of investing is fairly obvious (buy assets you believe to be undervalued or assets you expect to increase in price) just very hard to do consistently, portfolio construction is far more subtle. One of the issues that always causes a big debate is what is the right level of diversification for an investor.

Making this decision isn't helped by the fact that we get very different advice from different people. For example, Warren Buffet has called diversification ‘diworsification’ and suggested that most people be better off investing their money in their few best ideas. He introduced a concept of a 20 punch hole card representing the 20 investments one can make in one’s lifetime as a way of improving investors’ stock selection. At the other extreme ‘quants’ will own hundreds of stocks and index investors like Jack Bogle believe that the only correct level of diversification is owning the whole market. Even more confusing is that both sets of investors, and a lot of people in between, have good evidence to support their view. Given the breadth and strength of views that exist on the subject of diversification I think we often make the following key mistakes when thinking about the topic:
  • Viewing the right choice of diversification as independent of one's style of investing and fixed in time. i.e. what is right for me now is always right for me and must be right for you.
  • Ignoring risk in portfolio construction decisions.
I strongly believe that no one method fits all investors and the key to portfolio construction is getting a diversification level that matches your skill level, personality and asset selection criteria. Therefore rather than tell you that you should have a 5 or 50 or 500 asset portfolio I’m going to share some simple principles I believe that you should consider when deciding level of diversification is right for you.

Principle 1 - Diversification levels should increase as uncertainty increases.

This means that all things being equal beginners should have a higher diversification level than experienced investors who have a track record of being able to consistently identify mis-priced assets.

It also means that investors who do exhaustive research into the details of the company should have lower diversification levels than investors who rely on more quantitative approaches. If you believe that you have found genuine mis-pricings because you understand specific companies better than the market then it makes sense to hold enough of the shares of those companies to take full advantage of those mis-pricings. However if you rely heavily on heuristics like low P/E ratio or high Stockopedia StockRanks. then you capture the returns most consistently by being highly diversified. You can be confident that as a whole group low P/E stocks will outperform, but there is high uncertainty to which exact stocks in that group will drive that out-performance. Note that the research of Joel Greenblatt & Tobias Carlsisle shows that investors are actually really bad at choosing which subset of ‘Deep Value’ investments really outperform! See Carlisle’s Investors @google talk for details:

If you can find ways to reduce uncertainty by having greater control over the companies you invest in then you can hold more a more concentrated position. This is why it is rarely wrong for people like Buffett to have a significant proportion of their funds in companies they own outright where they can control the management and strategic direction of the business. As a pure investor if a larger stake reduces diversification but enables you to positively influence the management then this can be a good thing. However one of the mistakes pure investors often make is holding a large position in their current or former employer where they don’t have board level influence. This exposes them to the combined risk of employment and concentrated investment. Loyalty is not a good reason to be under-diversified, only superior knowledge or significant influence is.

Principle 2 - Diversification levels should increase as the consequences of making individual mistakes increases.

Diversification protects from an unexpected company or asset specific event having a large negative effect on one’s wealth. Therefore it makes sense not just to take account of how likely a negative event is to occur but the consequences of it occurring. I am far more careful when walking along a steep mountain ridge than walking across a field not because I regularly fall over in either just that the consequence of doing so are far more serious in the former. Likewise those who have large portfolios in comparison to their regular income should be more diversified not because they are more likely to be wrong just that the consequences of them being so are more severe.

This principle also leads one to consider liquidity in the decision-making process. As one’s portfolio get’s bigger the ability to take advantage of mis-pricings in smaller cap stocks may diminish since the risk of a large holding in an illiquid stock is that you can’t get out if something goes wrong. This factor may lead to higher diversification than would otherwise be chosen in experienced investors.
Also as one’s wealth increases additional wealth has a diminishing positive impact on one’s life, (what economists call diminishing marginal utility.) So even excellent stock-pickers may choose to diversify into other asset classes to reduce the risk that equities as an asset class suffer extreme negative returns as they get wealthier.

Note that this principle can be in competition with the first. When we begin our investing journeys we have much greater uncertainty in our ability which should lead to higher diversification but we also usually start with small sums relative to our future lifetime expected earnings which should lead to lower. With these decisions there is no one correct level but a balance of the unique factors that impact each of us.

Principle 3 – Err on the side of greater diversification

We all have behavioral biases. This means that you will almost certainly be overconfident in your ability to pick stocks, overconfident in the amount you truly know about each individual stock, and have hindsight bias remembering your past successes with concentrated investing and forgetting your mistakes. (Or at least engaging in attribution bias and blaming someone else for their failure as an investment!) All of these mean that you will underestimate the real level of uncertainty and underestimate the consequences of mistakes. This may lead to lower diversification levels than would be optimal. This doesn’t mean that everyone should hold a highly diversified portfolio just that when you consider these things you should very consciously err on the side of greater diversification not less.

So you can see that if like Buffet you are an experienced investor that has a strong track record of identifying mis-pricings through extensive analysis and taking influential stakes in those companies then high levels of diversification is diworsification.

However if you are a relatively new investor relying on some simple metrics to help you make investing decisions it probably wouldn’t harm you to add a couple more stocks to your portfolio.

In my next post on portfolio construction I’ll introduce a decision-making framework that can help you weight assets within your portfolio.

No comments:

Post a Comment